Assessing the impact of floods and federal policies on older adult mortality:leveraging Bayesian, quasi-experimental, and agent-based modelingtechniques

ABSTRACT Floods account for the majority of global climate change-attributable deaths in the 21st century and are predicted to increase substantially as the climate continues to rapidly change. As such, the health toll of floods stands to increase and will disproportionately affect the most vulnerable populations – particularly older adults. Older adults face heightened flood vulnerability due to disproportionate residence in floodplains, frailty, diminished hearing and vision, pre-existing chronic conditions, reliance on electricity-dependent medical equipment, and dependence on healthcare access. Thus, identifying policies to safeguard older adults' health during and following floods is paramount. In the US, the Federal Emergency Management Agency (FEMA) is tasked with prevention and mitigation of the effects of climate disasters. Two programs that are particularly important in the relief efforts from flood events are FEMA's National Flood Insurance Program (NFIP) and the Individuals and Households Program (IHP). These programs provide resources and services to individuals before, during, and after climate disasters – including major floods. As the rate of floods dramatically increases, understanding the ways in which floods affect mortality and these programs stand to buffer those effects is paramount. A critical research gap remains about (1) the effects of these FEMA programs on health, and (2) how long major floods may affect health beyond the initial days. This project will examine the effect of these programs on all-cause mortality among older adults, and will estimate both the short- and long-term effects of floods on all-cause, cardiorespiratory, and injury- related mortality at various time lags. Our overarching hypothesis is that floods affect health for up to a year following a major flood and that programs such as NFIP and IHP mitigate these effects. To test this hypothesis, we aim to (1) estimate the effect of NFIP availability on ZCTA-level mortality among Medicare enrollees after major floods and use an agent-based model to simulate NFIP mortality benefits, (2) estimate the effect of FEMA IHP application approval rate on mortality among Medicare enrollees after major flood exposure and use an agent-based model to simulate IHP mortality benefits, and (3) assess the relationship between exposure to major floods and all-cause, cardiorespiratory, and injury-related mortality at the county level in the short- and long-term. The proposed research includes a high-quality training plan that includes academic and scientific support for the success of the applicant in demographic methods, epidemiologic methods, causal inference and quasi-experimental design, climate epidemiology, Bayesian methods, agent-based modeling, and professional development. All of the proposed research will be supported by an interdisciplinary team and environment at the University of Washington. Collectively, the project will provide evidence that can inform future climate adaptation and FEMA policies to reduce preventable flood-attributable mortality.